Introduction
Head and neck cancers represent 10-15% of all cancers and cause 4-5% of all deaths from cancer. Annually, worldwide, approximately 650,000 patients are diagnosed with head and neck cancers, and 350,000 patients die from this disease(1-4).
Growing incidence and mortality of head and neck malignancies imposed therapeutic outcomes analysis, detecting the causes of post-therapeutic failures, and the study of genetical and molecular changes impacting the prognosis(1-4).
Materials and method
Statistical data necessary for processing sheets were collected from ONC1 initial declaration files, from ONC2 sheets of evolution and ONC3 death records present in the archives of the Medical Oncology Clinic of Craiova and Medical Oncology Cabinet County Craiova.
The overall incidence of head and neck cancers in Dolj county was 3.23‰ (524/17916), with an annual average of 20 cases, but an exponential increase in the number of patients registered for all neoplastic locations, during the study period.
Most patients included in the study were men (463 men/61 women, index ratio 7.59/1), elderly (median age 51.7 +/- 13.46 years), with provenance from the country side - rural/urban ratio: 1.81 (338/186 patients) -, dragged (chronic alcoholism, chronic smoking), but with acceptable performance status: 0-1 IP: 463 patients (88.35%) vs. IP 2: 61 patients (11.65%).
Most patients had carcinoma of the hypopharynx (210 patients; 40.07%) and larynx (195 patients; 37.21%) vs. oropharynx carcinoma (85 patients; 16.22%) and nasopharynx (34 patients; 6.48%). In the endoscopic confirmation of malignancy, it was detected the predominance of squamous carcinomas, moderately or poorly differentiated, characterized by marked tumor aggressiveness (493 patients; 94.08%). Patients randomized in the study presented as a whole and for each tumor location, locally advanced disease - stage III and IVA, B, respectively, clinical category T3 + T4 N1-3: 445 patients (84.92%) vs. stages I and II with clinical category correspondent T1 + T2 N0: 79 patients (15.07%). Localized disease stages were the prerogative of nasopharynx carcinoma (9/34 patients, 26%), carcinoma of the oropharynx and palatine tonsil more precisely (19/85 patients, 22.35%) and of the larynx glottis carcinomas (51/195 patients; 26.15%).
The primary objective of the study was to identify the most common errors in diagnosis and treatment responsible for therapeutic failures and detecting in this group of patients the expression of tumor markers that enable a correct therapeutic strategy after histopathological and immunohistochemical analysis.
Diagnostic errors
An error was practicing surgery as a first therapeutic gesture, that has been found in patients diagnosed with carcinoma of the oropharynx, hypopharynx and larynx with initial presentation to ENT clinics; after radical surgery, based on histopathological analysis bulletins, the patients were classified in stages III (pT3N1-3) or IVA, B (pT4N1-3) of the disease. Patients were randomized balanced in the two treatment arms, radiotherapy vs. radiochemotherapy, but all patients had therapeutic failures explained by an increased risk of distant dissemination, as a result of the practice of surgery as a first therapeutic gesture in locally advanced disease with lymphatic spread already present.
Another therapeutic error was reported in 16 patients with carcinoma of the oropharynx, larynx and hypopharynx (stage III cT3N0/1), which showed complete remission post-radiochemotherapy. Clinical and imaging lack of appreciation of reconversion to operability in the oncologic control and therefore lack of practicing surgery or irradiation dose supplementation were responsible for therapeutic failures in the further evolution of these patients.
In the comparative analysis of the therapeutic results, there were 524 patients included, clinically and radiologically diagnosed and histologically confirmed with carcinoma of the hypopharynx, larynx, oropharynx and nasopharynx at any stage of the disease, except metastatic disease (stage IVC), who were randomized balanced for all lots and all neoplastic locations into two therapeutic groups:
-
Group A included 254 patients treated using radiochemotherapy according to standard protocols: concurrent radiochemotherapy with cisplatin 20 mg/m²/day PIV CI x 5 days/4 weeks or 40 mg/m PIV CI/weekly or alternative radiochemotherapy using the regimen: 5-FU 1000 mg/m²/day i.v. day 1-4 CI + cisplatin 20 mg/m²/day i.v. on day 1/4 weeks.
-
Group B included 270 patients treated with Theratron 75, energy Mv 1.25 (range 2000-2006) or Varian linear accelerator with 6 MV energy (2007-2008); the total dose was 68-72 Gy/primary tumor (1.8-2 Gy / fraction, 5 days/week) and 40 Gy/laterocervical region (2 Gy/fraction, 5 days/week).
The rate of response was assessed at 2 months post-treatment. The remission rate was 59.92% (314/524 patients) but with a higher statistically significant percentage of complete responses (complete response: 188 patients, 35.87%, p=0.01). The threshold of statistical significance was reached on multivariate statistical analysis in the function of tumor location (p=0.053), disease stage (p=0.0001) and administered treatment (p<0.0001). Complete therapeutic response rate was statistically significant reduced in tumors of the hypopharynx and larynx, locally advanced, especially under the presence laterocervical lymph nodes metastasis (cN2,3), bulky, sores, infected, with tumor necrosis and hypoxia with radio-resistance, the therapeutic response being conditioned by category cN.
The disease-free interval
Patients who experienced complete remission and were recorded as failures benefited from reirradiation (according to TD administered initially and irradiated volume), salvage surgery (lymphadenectomy) or second-line palliative polychemotherapy for patients with locoregional evolving disease based on regimens with taxane (Paclitaxel/Docetaxel) and platinum (cisplatin/carboplatin) or first line palliative polychemotherapy for patients with locoregional or distant recurrences, 5FU + cisplatin regimen.
Thus, from a total of 188 patients (35.87%) with complete remission post-treatment, 95 patients (50.53%) experienced treatment failure after a median disease-free interval of 7 months. Most treatment failures were recorded as evolutionary locoregional disease, the consequence of therapeutic errors or of practicing surgery as a first gesture in advanced locoregional disease (73 patients; 76.84%), or lack of practicing surgery or irradiation dose supplementation in stage III with reconvertion to operability (16 patients; 16.84%). A small percentage of treatment failures were recorded as resuming locoregional evolution of the disease and is the consequence of diagnostic errors, with understaging of the disease, lack of advanced imaging and therapeutic practice of surgery as a first gesture (6 patients; (6.32%).
In patients receiving radiochemotherapy, from the 144 patients who had complete remission, 51 patients (68.62%) with carcinoma of the hypopharynx, larynx and oropharynx advanced locoregional experienced failures. In 35 patients who achieved therapeutic failures in terms of continuing the locoregional evolution of the disease, surgery was practiced as a first therapeutic gesture. 16 patients (31.37%) with treatment failures recorded in terms of resuming locoregional evolution of the disease after complete remission the therapeutical errors consisted in the lack of practicing surgery or at least supplementing the dose of irradiation. This aggressive development in terms of achieving complete remission after primary therapy raised the issue of performing immunohistochemical staining in order to detect molecular markers with prognostic value.
In the case of radiotreted patients, the 44 patients who experienced complete remission, therapeutic failures were the result of errors in diagnosis and treatment, being registered at a rate of 100%. Continuing the evolution of the disease in 38 patients with carcinoma of the hypopharynx, larynx and oropharynx in locally advanced stages was due to therapeutic errors, with the practice of surgery as the first therapeutic gesture. Resumption of locoregional evolution of the disease in 6 patients (4 patients with carcinoma of the larynx glottis and 2 patients with carcinoma of the nasopharynx) was consequently to diagnostic errors or failure to comply with therapeutic protocols or understaging.
On multivariate statistical analysis of the disease-free intervals, based on the tumor location and stage of disease, the threshold of statistical significance was achieved (p=0.0002, p=0.0091), demonstrating the prognostic value of tumor location and stage of the disease in terms of interval free of disease. In multivariate analysis of disease-free interval depending on the therapy administered, threshold did not reach statistical significance (p=0.3179); the failures were recorded in both arms after a median disease-free interval of 7 months, but with a rate of 3 times lower (35.41%) in patients receiving radiochemotherapy. Radiosensitising chemotherapeutic agents have a major role in increasing the effectiveness of radiotherapy and consequently the rate of therapeutic response, and improving locoregional disease control remains the objective of hyperfractionated radiotherapy.
Another objective of the trial was to analyze the survival rate at 5 years, being registered a 40.07% cancer mortality (210 deaths), with a median survival of 19 months.
On multivariate statistical analysis based on the location of neoplastic disease stage and treatment administered, the results reached statistical significance (p<0.0001, p<0.0001, respectively p<0.0001), demonstrating the prognostic value of the parameters studied, in terms of the rate of survival at 5 years.
Concurrent chemoradiotherapy or alternative demonstrated a statistically significant benefit in terms of survival at 5 years (median survival 18.8 months vs. 17.2 months, with a benefit of 15%, p<0.0001).
Our results are consistent with the results of meta-analysis of 87 randomized trials that included 16.640 patients, which demonstrated that concomitant radiochemotherapy leads to an 8% benefit for overall survival at 5 years, p <0.0001.
Multivariate analysis has detected the factors with predictive value in terms of response rate and factors with prognostic value in terms of survival rate at 5 years and the disease-free interval at 5 years, being the tumor location, disease stage and the treatment given. In the category including patients with poor prognosis are those diagnosed with carcinoma of the hypopharynx and larynx, in advanced locoregional stages radiotreated.
Morphological study was performed in 93 patients who were evaluated in both histological and immunohistochemical aspects for molecular markers (EGFR, p53 and Ki67) which overexpressed; according to data published in scientific journals, they are correlated with a poor prognosis.
Immunoreaction for EGFR was membranous (score 0 = absence of immunomarking to tumor cells; score 1-3 = immunomarking weak positive, score ≥4 = strongly positive immunomarking). Nuclear immunomarking for proliferation index Ki67 was semiquantitative (score 1 - index ≤15%; score 2 - index of 16-44%, score 3 - index ≥45%). Nuclear immunomarking for p53 protein was semi-quantitative (score 0 = no; score 1 = nuclear immunomarking in less than or equal to 10% of tumor cells; score 2 = nuclear immunomarkingin in 11-54% of tumor cells; score 3 = immunomarking positive in at least 55% of tumor cells, and scores of 0 and 1 were considered negative for the presence of mutations in the p53 gene, and scores 2 and 3 positive for this mutation).
Immunohistochemical analysis detected linear relationship between the degree of tumor differentiation and membrane EGFR expression or p53 and Ki67 nuclear immunomarking. In epidermoid carcinomas poorly differentiated in most cases it was detected the overexpression of EGFR membrane (19/36 cases - 52%), nuclear immunomarking p53 in >55% of tumor cells (27 cases) and nuclear immunomarking Ki67 in> 45% of tumor cells (24/33 cases, 772.7250).
It was also noted a linear relationship between the expressions of tumor markers, carcinomas expressing intense membrane for EGFR presented an intense nuclear immunomarking for p53 (more than 55% of tumor cells) and Ki67 (more than 45%). So, patients with carcinoma who experienced a positive expression of EGFR (66 cases, 79.97%) had a Ki67 proliferation index greater than 45% (33 cases, 35, 48%) and p53 protein showed a positive nuclear immunomarking in cells in over 55% of the 27 cases (23.66%).
So, after immunohistochemical analysis we established a linear relationship between histopathologic type, degree of tumor differentiation and tumor marker expression.
The expression of molecular markers was detected immunohistochemical in patients receiving radiochemotherapy where the results reached statistical significance threshold in terms of therapeutic response rate, the overall survival and disease-free interval. The group with poor prognosis included 35 patients who were recorded as failures, with evolution of the disease, due to surgery practiced as a first therapeutic gesture in advanced stages of the disease and 16 patients with the resumption of the disease in the absence of practicing surgery after complete remission after therapy; death in this group was recorded when reporting data due to disease distant progression(23 cases of bone metastasis and 12 cases of lung metastases) and resumption of locoregional evolution of the disease (16 patients). Immunohistochemical staining detected intensely positive immunoreactions for EGFR membrane with nuclear immunomarking for p53 in a proportion of more than 55% in tumor cells and nuclear immunomarking for proliferative index Ki-67 in a proportion of more than 45% of tumor cells.
The group with intermediate prognosis included 38 patients out of 53 patients with stationary disease after radiochemotherapy, with resumption of disease progression after a median free interval of 7 months, the death being registered in reporting data by resuming locoregional evolution. Immunohistochemical stains detected weak positive immune reaction for EGFR with nuclear immunomarking for p53 in 10-55% of tumor cells and nuclear immunomarking for Ki67 in 10-45% of tumor cells.
The group with favorable prognosis included 93 patients of the 144 patients with complete remission after radiochemotherapy and who had a locoregional controlled disease when reporting the data. Immunohistochemical study due to material reasons was conducted only in 4 patients with poorly differentiated epidermoid carcinoma of the larynx. Immunohistochemical staining detected the absence of membrane expression of EGFR and p53 nuclear immunomarking was absent or present in <10% of tumor cells (0/1 score correlated with the absence of a mutant form of p53) and Ki-67 proliferative index was absent or present in <10% of tumor cells.
Following immunohistochemical study, the immunophenotypic groups could be defined as follows:
-
a. favorable prognosis: absence of immunomarking for EGFR, nuclear immunomarking for p53, Ki67 absent or present in <10% of tumor cells;
-
b. intermediate prognosis: immunomarking weakly positive for EGFR, p53 nuclear immunomarking in 10-55% of tumor cells and nuclear Ki67 immunomarking in 10-45% of tumor cells;
-
c. unfavorable prognosis: immunomarking strongly positive for EGFR, p53 nuclear immunomarking positive in >55% of tumor cells and nuclear immunomarking Ki67 positive in >45% of tumor cells.
The contribution of the authors in this vast tumoral pathology is apparent from the detection of clinical and immuno-histoprognostic factors (tumor location, disease stage, treatment administered, histopathological type, degree of tumor differentiation, the expression of tumor markers) with the possibility of a correct therapeutic randomization based on histopathological and immunohistochemical results. It also requires a correct diagnosis when randomising patients based on ENT examination in combination with performant imaging to avoid surgery practice as a first gesture in locally advanced stages, a therapeutic error which is a major a cause of failures. The group with poor prognosis included patients diagnosed with poorly differentiated carcinoma of the hypopharynx and larynx in locally advanced stages with intensely positive membrane overexpression of EGFR and strongly positive nuclear immunomarking for p53 and Ki67, who required concomitant radiochemotherapy.
Discussions
Choosing the appropriate treatment remains a controversial topic, considering the poor prognosis of locoregional advanced stages with the presence of tumor formations over 3 cm in size and/or laterocervical lymph nodes metastasis, with a curability rate of 30%(5-7). According to the results of “MACH-NC meta-analysis of Chemotherapy in Head and Neck Cancer”, based on 15 randomized trials involving 2487 patients, multi-agent induction (regimes platinum + 5FU) + radio-chemotherapy vs. exclusive radiotherapy lead to a benefit of 5% in overall survival at 5 years (p=0.05)(8-17). The results of a meta-analysis of 87 RCTs that included 16,640 patients showed that concurrent radiochemotherapy with Cisplatin vs. radiotherapy alone determines an exclusive benefit of 8% in overall survival at 5 years (p<0.0001). So, multi-agent induction followed by concurrent radiochemotherapy is the standard treatment in locally advanced stages and induction polychemotherapy regimens with taxanes (especially Paclitaxel) and platinum are cited in literature as the most effective, with rates of therapeutic responses of 71% to 100% with a 1-year survival rate of 69-98% and a 3-year survival rate of 41-82%(18-21).
The results of Bonner’s study showed that the combination of Cetuximab (Erbitux) to radiotherapy determines an absolute benefit for survival (10% at 5 years), similar to that achieved by the addition of other chemotherapy regimens to radiotherapy (8% at 5 years) and represents an alternative to the clasical radiochemotherapy with similar efficacy and acceptable toxicity for locally advanced head and neck cancers with increased risk of local recurrence and/or locoregional (cT2-4 and cN2,3)(24-26).
A new approach to radio technology is required to improve locoregional control of the disease, namely, accelerated HART (Hyperfractionated radiotherapy), intensity modulated radiotherapy correlated with dynamic contrast imaging magnetic resonance (imaging-guided radiotherapy - GRT)(27-30).
According to data published in scientific journals, hyperexpression of p53 is correlated with the presence of mutant p53 form at a score of 1-2 (immunomarking present in 10-50% of tumor cells), expression of Ki67 index correlates with mitotic activity and EGFR hyperactivity indicates an uncontrolled division of cells and is present in 90% of head and neck epidermoid carcinomas(22-23).
Immunohistochemical study correlated with clinical study helped us to define the clinical and immunohisto-prognostic groups as follows:
-
a. favorable prognosis: absence of immunomarking for EGFR, nuclear immunomarking for p53, Ki67 absent or present in <10% of tumor cells;
-
b. intermediate prognosis: immunomarking weakly positive for EGFR, p53 nuclear immunomarking in 10-55% of tumor cells and nuclear Ki67 immunomarking in 10-45% of tumor cells;
-
c. unfavorable prognosis immunomarking strongly positive for EGFR, p53 nuclear immunomarking positive in >55% of tumor cells and nuclear immunomarking Ki67 positive in >45% of tumor cells.
The study results demonstrate and emphasize the importance of establishing a correct diagnosis based on clinical ENT examination correlated with performant imaging techniques (CT scan/head and neck MRI with contrast material), avoiding understaging and implicitly therapeutic errors consisting in practicing surgery as a first therapeutic gesture in locoregional advanced stages, therapeutic errors responsible for locoregional failures or distant disease. The key to a successful treatment is the possibility of a correct therapeutic randomization based on histopathological and immunohistochemical analysis, and the group of patients with poor prognostic is required the standardization of radiation therapy and anti-EGFR association.
Conclusions
Correct staging is based on clinical ENT examination correlated with imaging examinations for evaluating the real extension of the disease and thus reducing the risk of errors and of therapeutic failures, due mainly to practicing surgery as a first therapeutic gesture in advanced stages of the disease.
The group with poor prognosis includes patients diagnosed with squamous cell carcinoma poorly differentiated of the hypopharynx and larynx, in advanced locoregional stages, that in the absence of performant imaging techniques were understaged; in consequence, surgery was practiced as a first therapeutic gesture, a therapeutic error responsible for treatment failures recorded in terms of distant progression of the disease (lung metastasis and/or bone). In this group of patients showing hyperexpression of tumor markers (EGFR, p53 and Ki67) it is required that molecular targeted therapy should be administrated concurrent with radiotherapy.
Clinical phenotypes with predictive and prognostic value in ENT cancers are tumor location, disease stage and treatment given. Radiosensitising agents associated is the key to reducing 3 times the rate of local recurrence, but improvement of locoregional control remains an objective for performance-conformational radiation therapy with hyperfractionated dose regimen.
Immunohistochemical study correlated with clinical study helped us to define the clinical and immunohisto-prognostic groups based on the analysis of histopathological and immunohistochemical results, so that adequate therapy could be ensured.
In the pre-therapeutic balance is required a clinical and imaging diagnosis correlated with histopathological and immunohistochemical analysis, bulletins elements underlying a fair treatment to improve the prognosis of patients with risk and thus reducing therapeutic failures.
The group with poor prognosis includes patients diagnosed with squamous cell carcinoma poorly differentiated of the hypopharynx and larynx in advanced locoregional stages with strongly positive immunomarking for EGFR, nuclear immunomarking for p53 in more than 55% of tumor cells, and nuclear immunomarking for Ki67 positive in more than 45% of the tumor cells; that requires the association of Erbitux treatment to standard induction therapy and/or radiotherapy. n